Team:Tec-Monterrey/Hardware

HARDWARE

Microfluidics

Microfluidics is a subset of fluid mechanics that integrates physics and engineering on a micrometer length scale [2]. For microfluidic systems, flow rates operate in small volumes, ranging from microliters (10^-6) to picolitres (10^-12) inside micrometric (10^-6 μm) channels. The small dimensions in which these systems are driven allows a wide number of advantages over conventional techniques that use reaction vessels, test tubes or microliter plates, which include more predictable systems with high-throughput experimentation capabilities, a smaller volume of sample needed, a shorter time to gather results, and the possibility to execute multiple parallel tests [5], [15].

Droplet microfluidics

Microfluidic systems take advantage of the properties of fluid flow on the microscale to the widest possible range. Research in this area include theoretical analysis, fluid behavior and applications based on this regime [16]. Microfluidic platforms operate both in the continuous flow regime and in the segmented flow regime. In the second case, droplets are used as individual reaction chambers, alike conventional techniques used in standard laboratories, but orders of magnitude smaller with higher control and precision [3].

Droplet-based microfluidics is an adaptable tool for extensive applications, attributed to have a number of advantages [6],[8]-[10]. In a channel microfluidic device, two separate and immiscible liquid streams are forced into channels. The shear force at the point where they meet creates (usually aqueous) droplets that are encased within a continuous fluid (generally oil). Each droplet is a self-contained vessel, isolated from other droplets and from the channel walls by the oil phase.

Droplet Generation Techniques

Several different techniques for droplet generation have been extensively proposed and reviewed in macroscopic configurations [1],[12],[15]. Nevertheless, these methods cannot produce monodisperse droplets, thus making them unsuitable for biological or medical applications. Hence, microfluidic techniques have been proposed in order to overcome these limitations. The highly complex physics of droplet generation in microfluidic systems is not completely understood yet, but there is enough information in order to implement droplets on microfluidic devices successfully. There are three main techniques for droplet generation that have been conventionally used: T-junction, co flow, and flow focusing. Each technique has a different mechanism of pinch-off in order to form the droplets as depicted in figure 1.

Figure 1. Schematic representation of the most popular droplet generation methods [17].

T-junction

The T-junction method uses a simple junction (often formed like the letter “T”) where the continuous and the dispersed phase meet at an angle. The disperse phase upon reaching the continuous phase pinches off in the droplet form, induced by shear stress exerted by the continuous phase, and gets carried in its stream [13]. The droplet detaches from the fluid thread before it can block the full width of the continuous phase channel. This is a straightforward method and different regimes are obtained based on the capillary number [9].

Co flow

In the co flow design, the dispersed phase and the continuous phase meet in parallel. A capillary is inserted in a surrounding channel in such a way that the fluid in the surrounding channel completely engulfs the capillary and any liquid that is ejected out of it [12]. Therefore, the channel carrying the water phase is completely covered by the oil stream. This method is again dominated by the capillary number or rather the flow rates of the two streams yielding different flow regimes [4].

Flow focusing

This model is a modified version of the co-flow device and produces more uniform droplets than the T-junction and the co-flow systems. Here the co flowing of two or more phases are accelerated and elongated in a constriction of the channel (orifice). The presence of the orifice structure in the parallel flow system creates different flow regimes based on the flow rates of the two phases [13].

Comparison of methods for droplet generation

The techniques mentioned above are used to produce droplets in numerous different purposes. The geometry used is dependent on the application and the specifications needed. The droplet size depends on several parameters such as the geometry, the flow rates and the liquids being used. Flow focusing devices usually need a more precisely controlled production and accurate control of parameters in order to produce droplets at a higher monodispersity and can generate smaller size droplets at a higher rate compared to T-junctions. Co flow devices’ behavior is hard to control compared to T-junctions or flow focusing and finally T-junctions may not generate droplets as monodisperse as the ones in a flow focusing device, however this kind of device is the most used and documented.

As these methods require a precise control of the flow rate to generate monodisperse droplets, which is difficult to reach especially with syringe pumps, but taking into account our resources and time availability we decided to go ahead with a T-junction device, especially because we found many references that define parameters we can use instead of obtaining them ourselves.

Microfluidic device design

In order to fulfill the design considerations regarding the microfluidic device, a number of different design ideas were sketched using AutoCAD 2019. Once a microfluidic device design was sketched, it went through a redesign and development process in order to have an optimized model based on previous literature and known good practice parameters. As previously discussed, the microfluidic device should have a droplet generator, a spacer, a second droplet generation, a reservoir, and an outlet for recovery. For the droplet generator, two geometries were taken into account, a T-junction and a flow-focusing device, but based on a literature search and the microfluidics model section, a T-junction was preferred based on its simplicity. The following figure shows the first sketched models.

Figure 2. Different proposed design for our device.

A microfluidic device was designed based on the microfluidics model parameters section in order to generate monodisperse droplets in two different junctions. The first junction needed to generate droplets using a T-junction, then go through a spacer in order to prevent coalescence or merging of the droplets, then a second T-junction would be added in order to generate the second droplet, after these droplets were mixed, a reservoir chamber is used to visualize and count the droplets, and finally an outlet for recollecting the mixed droplets. The final proposed design is shown in the following figure.

Figure 3. Detailed final design of our device.

Microfluidic device fabrication

Once the design was validated as a good candidate for the requirements given, a number of different variations were also made by adding a second spacer after the mixing, and changing the number of spacers. These different versions were added into a single file and a mask was ordered from CAD/Art services, inc.

Figure 4. Variations of design included in our final masks.

With the masks, a master mold was fabricated using SU8-2007 over a glass slide. Although a silicon wafers is the gold standard to use as a substrate, we couldn’t use them because they are too expensive. This process was based on the MicroChem SU-8 data sheet as a starting point, but we had to make a number of changes and develop our own process based on the availability of resources.

One of the biggest issues was that we didn’t have a cleanroom, hence all fabrication procedures were made in a dark room, however dust and other particles could be in the environment, making the process more difficult with the possibility of having a non-functional master mold or blocked channels. After the fabrication of the SU8 master, a microfluidic device with the proposed geometry was fabricated using standard soft lithography techniques.

The polydimethylsiloxane (PDMS) devices were bonded onto a glass slide and sealed to form the microfluidic channels using a plasma surface activation. Once the final microfluidic device was made, the proposed flow conditions were tested and the results were compared with the simulations (see model). Syringe pumps were used to maintain a continuous flow rate of both the disperse and continuous phases, the syringe pump flow rate is defined by the user and is the main control parameter for the generation of the droplets.

Experimentation

A set of conditions based on the droplet breakup simulations were taken into account, the flow rate of the continuous phase was changed from 1 to 50 µm/min, while leaving the disperse phase constant at 1 µm/min in order to use the smallest possible amount of reagent to optimize the biological sample. Two droplets of different content are forced to be in contact and to merge, then the mixed droplet passes through the remaining channel into a reservoir to contabilize and see the monodispersity of the generated droplets. Finally, the droplets are recaptured in a vial for further experimentation.

To see further details and an extensive explanation of the fabrication process please see our protocols document.

References

1. Anna, Shelley Lynn. (2016). “Droplets and Bubbles in Microfluidic Devices.” Annual Review of Fluid Mechanics 48: 285–309.
2. Baroud, Charles N., Francois Gallaire, and Rémi Dangla. (2010). “Dynamics of Microfluidic Droplets.” Lab on a Chip 10(16): 2032–45.
3. Debon, Aaron P., Robert C.R. Wootton, and Katherine S. Elvira. (2015). “Droplet Confinement and Leakage: Causes, Underlying Effects, and Amelioration Strategies.” Biomicrofluidics 9(2).
4. Ganesh, Shruthi. (2017). “UC Irvine UC Irvine Electronic Theses and Dissertations.” : 123.
5. Gu, Hao, Michel H.G. Duits, and Frieder Mugele. (2011). “Droplets Formation and Merging in Two-Phase Flow Microfluidics.” International journal of molecular sciences 12(4): 2572–97.
6. Jeong, Woong-Chan et al. (2012). “Controlled Generation of Submicron Emulsion Droplets via Highly Stable Tip-Streaming Mode in Microfluidic Devices.” Lab on a Chip 12(8): 1446–53.
7. Johnson, Paul Michael. 2015. “UC Irvine UC Irvine Electronic Theses and Dissertations.” : 123.
8. Kang, Zhanxiao et al. (2016). “Engineering Particle Morphology with Microfluidic Droplets.” Journal of Micromechanics and Microengineering 26(7): 75011.
9. Kong, Tiantian et al. (2013). “Engineering Polymeric Composite Particles by Emulsion-Templating: Thermodynamics versus Kinetics.” Soft Matter 9(41): 9780–84.
10. Kong, Tiantian et al. (2012). “Droplet Based Microfluidic Fabrication of Designer Microparticles for Encapsulation Applications.” Biomicrofluidics 6(3): 34104.
11. Martz, Thomas D et al. (2012). “Microfluidic Generation of Acoustically Active Nanodroplets.” Small 8(12): 1876–79.
12. Rosenfeld, Liat, Tiras Lin, Ratmir Derda, and Sindy K Y Tang. (2014). “Review and Analysis of Performance Metrics of Droplet Microfluidics Systems.” Microfluidics and nanofluidics 16(5): 921–39.
13. Schuler, Friedrich. (2016). “Centrifugal Step Emulsification for Digital Droplet Amplification.”
14. Tarn, M.D., and N. Pamme. 2014. “Microfluidics.” Reference Module in Chemistry, Molecular Sciences and Chemical Engineering: 1–7.
15. Vladisavljević, G T, Isao Kobayashi, and Mitsutoshi Nakajima. (2012). “Production of Uniform Droplets Using Membrane, Microchannel and Microfluidic Emulsification Devices.” Microfluidics and nanofluidics 13(1): 151–78.
16. Zhu, Pingan, and Liqiu Wang. (2017). “Passive and Active Droplet Generation with Microfluidics: A Review.” Lab on a Chip 17(1): 34–75.